I want to measure the absolute humidity in a gas stream, and i've found that water has an absorption peak at ~940-950nm, which is really nice since it means that one just might be able to use an IR diode from a remote control as radiation source, and a corresponding photodiode for detection. These diodes seem to have an emission spectrum that very conveniently overlaps with the water absorption.
Since i'm a chemist and not an EE, i come to you for some advice on how to actually set this up. Sadly, the absorption isn't very strong (and i don't want to use a long pathlength for a number of reasons), so i will have a large signal - beam intensity - on which i want to measure a small change - the attenuation. I read somewhere that using the photodiode without any voltage connected to it and measuring its output current when illuminated would result in the least noise, is this correct? This means i have to convert the current to a voltage to measure with the arduino's AD converter, which is done using a transimpedance amplifier (basically an ordinary amplifier with its output connected to one of its inputs via a resistor/cap in parallel). Likely, i'll also need to amplify and scale this output voltage to -hopefully- be able to see the minute changes in attenuation.
If i've got something wrong or if there's a better way to do this, please tell me! If the above is correct, which kind of amplifier should i use, an instrumentation amplifier? I've noticed that they can be quite expensive, and from what i can understand most distortions/noise is the result of rapid changes in signal (correct?), so since my signal will be more or less constant and i only want to sample at perhaps 20Hz i might aswell use an ordinary, cheap op-amp? Could/should i use the same amp for both stages?
Normally, you'd want two detectors - one monitoring the source directly to compensate for variations in intensity, and the second monitoring the modified (absorbed) signal.
That would add a collimating lens, beamsplitter (half-reflective mirror), full-reflective mirror and another photodiode to the setup. Nothing expensive, but all this added complexity just to get rid of fluctuations in the lightsource? I've never constructed an optical system like this before, sure would be fun to try, but...if the lightsource fluctuations are reasonably faster than the sampling rate (design goal 20 Hz), then wouldn't one get the same result just by averaging/filtering those out?
If i tried this out, then should i connect the transimpedance amplifier inputs to the two photodiodes? Would that create a voltage output that's proportional to the difference between the currents from the two photodiodes? Or first convert each photodiode current to voltage, then amplify the difference between these?
And can i use just about any op-amp or is there some type that would be more suitable? The -amount- of information about this out there is just...overwhelming.
That would add a collimating lens, beamsplitter (half-reflective mirror), full-reflective mirror and another photodiode to the setup
Not necessarily - you could simply point the reference detector at the source, as long as you can null-out the absolute difference.
It probably doesn't even matter if it is off-axis of the main beam for the same reason. (a similar system used to be used on early x-ray tomography machines)
Besides, all those things are salvageable from a CD drive
but...if the lightsource fluctuations are reasonably faster than the sampling rate (design goal 20 Hz), then wouldn't one get the same result just by averaging/filtering those out?
It may be - it depends on the nature of the readings and of the fluctuations.
I would use a trial and error approach increasing complexity only whenever is necessary.
First set up a simple system using a photo transistor as a detector (if you can find one sensitive to the wavelength you are using). They are normally more sensitive than photodiodes. If not, then try the photodiode.
-I would bias the diode or phototransistor and try to read the voltage output. You can set up a Wheastone Bridge configuration with a fix voltage divider on one side and the biased photodivice on the other. That allows you to cancel the bias voltage by using a differential amplifier without the need for an coupling AC capacitor. To set up your Diff Amp you don't need a special OpAmp for that trial as the frequency is low, the slew rate is low and the output impedance of the configuration is also low. You can try to use a regular LM324. In the datasheet you will find the Instrumentation Amp you can build with it and how to calculate the gain, etc.
-With that rig already built, perform some tests and find out what you get at the output when you change the variable. From the results you will get enough info to make the necessary corrections to achive what you want in terms of sensitivity, etc.
olof.gross:
I want to measure the absolute humidity in a gas stream, and i've found that water has an absorption peak at ~940-950nm
That peak is very weak in gaseous water, see Martin Chaplin's water pages on http://www.lsbu.ac.uk/water/vibrat.html
Even all the molecules in a whole atmosphere column reduce the solar intensity by only one half.
